Browse Prior Art Database

Autonomic parallel processing software architecture in distributed computing nodes Disclosure Number: IPCOM000032938D
Original Publication Date: 2004-Nov-19
Included in the Prior Art Database: 2004-Nov-19
Document File: 2 page(s) / 15K

Publishing Venue



Disclosed is software architecture of parallel computing for distributed autonomic processing nodes. The architecture provides an efficient distributed parallel processing of multiple independent parts which are divided from a large Job, and it let distributed nodes pull out the parts autonomously for their processing. The logic uses pull strategy by distributed autonomic nodes instead of push strategy by an application server to dispatch application job parts. The autonomic pull strategy reduces complexity of job dispatching and resource management. It also increases flexibility and availability of the computing system.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Autonomic parallel processing software architecture in distributed computing nodes

1.Distributed computing nodes
(1)The computing system consists of multiple autonomic server nodes.
(2)A Job management server manages claim checks for application job parts. And it manages execution status of jobs and job parts.
(3)Database servers manage databases required for applications.
(4)Application servers execute job parts.

2.Autonomic pull strategy for Job parts dispatching
(1)Application servers acquire the next available part of Job autonomously.
(2)Application servers judge whether they can execute the available Job part autonomously with their own resources.
(3)The Job management server does not need Job dispatching process and logics.
(4)The Job management server does not need complex error detection and recovery.

3.Claim check for load balancing and application server management
(1)The Job management server prepares the claim checks corresponding to each Job parts, and set processing ID, function, and parameters to the each claim check entries in the Job list table.
(2)The Job management server manages processing sequencing of each Job parts for load balancing of application servers in case each Job parts have different processing loads.

4.Autonomic parallel processing of independent Job parts
(1)Divide a large application Job into multiple independent parts.
(2)Register the parts to the Job List Table of the Job management server
(3)Initiate Job parts execution process of the appl...